The Fas/FasL system mediates induced apoptosis of immature thymocytes and peripheral T lymphocytes, but little is known about its implication in genetic susceptibility to T-cell malignancies. In this article, we report that the expression of FasL increases early in all mice after γ-radiation treatments, maintaining such high levels for a long time in mice that resisted tumor induction. However, its expression is practically absent in T-cell lymphoblastic lymphomas. Interestingly, there exist significant differences in the level of expression between two mice strains exhibiting extremely distinct susceptibilities that can be attributed to promoter functional polymorphisms. In addition, several functional nucleotide changes in the coding sequences of both Fas and FasL genes significantly affect their biological activity. These results lead us to propose that germ-line functional polymorphisms affecting either the levels of expression or the biological activity of both Fas and FasL genes could be contributing to the genetic risk to develop T-cell lymphoblastic lymphomas and support the use of radiotherapy as an adequate procedure to choose in the treatment of T-cell malignancies. [Cancer Res 2007;67(11):5107–16]

The Fas/FasL system provides the cell with an effective way of self-regulation through apoptosis. Its most remarkable role is played through the immune system, where the development of T cells within the thymus and later peripheral homeostasis constitutes critical events controlled, at least in part, by Fas/FasL-mediated apoptosis (1, 2). However, this system is also known to transduce proliferative or activating signals (3, 4). Consistent with these functions, point mutations in the Fas/FasL system have been implicated in some diseases of the immune system as well as in the pathogenesis of many tumor types (57).

Although the involvement of the Fas/FasL system in the apoptosis of peripheral T cells has been well established, little is known at present about its role within the thymus (8). Some studies suggested a role for Fas/FasL in thymocyte development, evidencing its involvement in negative and/or positive selection of thymocytes during early thymic development in gestation and in negative selection of thymocytes in adult mice depending on the dose of the antigen (912). Recently, Schmitz et al. have shown that mouse FasL can mediate T-cell receptor-induced apoptosis of CD4+CD8+ double-positive thymic lymphoma.

γ-Radiation is a prototypical carcinogen that acts by inducing DNA damage and subsequently cell death in radiosensitive tissues. Exposure of mice to moderate doses of whole-body irradiation may induce thymic lymphomas, particularly T-cell lymphoblastic lymphomas (13). Interestingly, stress from ionizing radiation exposure has been shown to lead to Fas/FasL activation and subsequent apoptosis in peripheral T cells (14, 15). Along this same line, we have shown that γ-radiation–induced mouse T-cell lymphoblastic lymphomas expressed lower levels of Fas mRNA compared with those found in control thymuses (16, 17).

Using a panel of interspecific chromosome substitution strains between SEG/Pas (Mus spretus) and C57BL/6J (Mus musculus), we further identified a thymic lymphoma resistance locus (Tlyr1) on chromosome 19 that included Fas as a putative candidate gene for tumor resistance (18). Interestingly, we speculated that some functional polymorphisms at the Fas promoter of inbred strains could be involved in their different susceptibility for developing such tumors (17). Additional polymorphisms at the coding sequence of FasL have also been described as affecting the biological activity of this protein (19).

Taken together, these findings suggest that the Fas/FasL system could be an important mediator in apoptosis of immature T cells, particularly following ionizing radiation, and that functional polymorphisms in both genes could be modulating the genetic risk to develop T-cell lymphoblastic lymphomas. Because we had already reported functional promoter polymorphisms of Fas between resistant and susceptible mouse strains (17), the goals of this study were as follows: (a) to determine if there existed differences in the levels of transcriptional expression of FasL in untreated thymuses and T-cell lymphoblastic lymphomas induced in mouse strains with extreme differences in susceptibility; (b) if so, to identify any changes in the nucleotide sequences of their promoter regions that could be responsible for their differential expression; and (c) to design and carry out functional analyses, both in vitro and in vivo, to determine the activity of the Fas/FasL system from each of the two strains.

Our results lead us to propose a role for both genes in contributing to the specific genetic susceptibility to T-cell lymphoblastic lymphomas.

Mice and treatments. C57BL/6J mice were obtained from The Jackson Laboratory. SEG/Pas was provided by Drs. J.L. Guenet and X. Montagutelli (Institute Pasteur, Paris, France). At least 115 mice from each strain were used for the transcriptional analysis: (a) 5 control nonirradiated mice from each strain, (b) 10 mice subjected to a single lethal dose of 10 Gy and sacrificed 24 h later, (c) up to 10 that had developed thymic lymphoma during the 25-week latency period after receiving four sublethal doses of 1.75 Gy on a weekly basis, and (d) 10 individuals not having developed lymphoma after this treatment were taken. Thymic lymphomas (in all cases T-cell lymphoblastic lymphomas) were characterized as described previously (18). In our experimental conditions, the strain C57BL/6J (derived from M. musculus) exhibited a global incidence of 73.4% (20). In contrast, SEG/Pas, derived from M. spretus, is extremely resistant with a tumor incidence of only 3% (18).

In addition, nine animals from each strain were used to do terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assays and/or Western blot (three untreated, three 1.75 Gy treated, and three 10 Gy treated).

For animal experimentation, we followed the ethical considerations dictated by the European Commission (Directive 86/609/CEE) for the humane care and use of laboratory animals.

Quantitative reverse transcription-PCR. RNA from the thymus was obtained using TriPure Reagent (Roche). The determination of the transcript levels was done by quantification assays based on real-time reverse transcription-PCR (RT-PCR) with a LightCycler instrument (Roche Diagnostics) using fluorescence resonance energy transfer (FRET) hybridization methodology.

The gene encoding the tubulin β5 chain (Tubb5) was used as an internal control of the RNA quality and amplification. Primers and FRET probes (designed by TibMolBiol) were as follows: Tnfsf6-S, 5′-TGGAATGGGATTAGGAATGTAT-3′; Tnfsf6-R, 5′-GTGTACTGGGGTTGGCTATT-3′; Tnfsf6-FL, 5′-TGAACTCACGGAGTTCTGCCAG-FL-3′; Tnfsf6-LC, 5′-LC Red640-TCCTTCTGCAGGTGGAAGAGCTG-PH-3′; Tubb5-F, 5′-TGGGACTATGGACTCCGTTC-3′; Tubb5-A, 5′-AAAGCCTTGCAGGCAATCA-3′; Tubb5-FL, 5′-GGCCTTTAGCCCAGTTGTTGCCT-FL-3′; and Tubb5-LC, 5′-LC Red705-CCCCAGACTGACCGAAAACGAAGTT-PH-3′.

Reactions for real-time PCR (using a one-step LightCycler kit) and subsequent coamplification of FasL and Tubb5 were done as described before (17). All PCRs were repeated for each sample in three independent experiments. FasL expression levels were calculated with the LightCycler Relative Quantification software (Roche Diagnostics).

Genotyping of the mouse FasL promoters. DNA from the thymus was extracted using DNAzol reagent (Molecular Research Center, Inc.).

Based on the sequence reported for the M. musculus FasL promoter in 2000 (Genbank accession no. AF045739; ref. 21), a couple of primers containing two distinct restriction sites for KpnI and XhoI were designed: FasLPromK-F, 5′-GGTACCCAAACTCTTCTGAATTAGGCA-3′ and FasLPromX-R, 5′-CTCGAGTAATTCATGGGCTGCTGCAT-3′.

Using the Expand High Fidelity PCR System (Roche Diagnostics) and the aforementioned primers, the region comprising from −706 to +20 was amplified. PCR products were sequenced. All sequencing reactions were done on an ABI Prism 310 Automated Sequencer (Applied Biosystems). All comparisons between sequences were made using the program L-Align from ExPASy Molecular Biology Server.

Cloning of the FasL promoters and reporter constructs. Purified DNA fragments containing the FasL promoters were ligated overnight at 4°C to pGEM-T Easy Vector (Promega) using the 2X Rapid Ligation Buffer according to their instructions. Using JM109 High Efficiency Competent Cells (Promega), transformation reactions were carried.

Fragments doubly digested with KpnI and XhoI (Roche) were cloned into luciferase pGL2-Basic plasmid (Promega).

Genotyping of the mouse Fas and FasL cDNAs. Total RNA from either liver (Fas) or testis (FasL) of C57BL/6J and SEG/Pas mice was extracted using TriPure Reagent.

Based on previously reported sequences for both Fas and FasL cDNAs (Genbank accession nos. NM_007987 and NM_010177, respectively), a couple of primers containing two distinct restriction sites for HindIII and BamHI (Fas) and MluI and NheI (FasL) were designed: FasHind-F, 5′-GCAAGCTTTTTTCCCTTGCTGCAGACATG-3′; FasBam-R, 5′-GCTGGATCCGAGGTAGTTTTCACTCCAGAC-3′; FasLMlu-F, 5′-AGTACGCGTCAGAGTTCTGTCCTTGACACC-3′; and FasLNhe-R, 5′-GATGCTAGCATTAAGGACCACTCCATGGACC-3′.

Reverse transcription was done using SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen) followed by PCR using the Expand High Fidelity PCR System and the aforementioned primers.

Cloning of the Fas and FasL cDNA sequences. Purified DNA fragments containing the Fas cDNA were doubly digested with HindIII and BamHI (Roche) and cloned into pcDNA3 plasmid (Invitrogen).

Purified DNA fragments containing the FasL cDNA were cloned into pGEM-T Easy Vector as described previously. Fragments doubly digested with MluI and NheI (Roche) were initially cloned into pBI-EGFP vector (Clontech Laboratories, Inc.) and then subcloned through BamHI and EcoRV (Roche) digestion into pcDNA3 plasmid.

Cell cultures and transient transfections. Human Jurkat T cells and human HEK-293 cells were cultured as described elsewhere (17, 22).

Transfections were done using LipofectAMINE 2000 Reagent for Jurkat and LipofectAMINE Reagent for HEK-293 (Invitrogen).

Luciferase assays.FasL promoter activity was determined through luciferase assays as described previously (17). Firefly and Renilla luciferase activities were monitored using a Sirius luminometer (Berthold).

Caspase-Glo 8 assays. Human HEK-293 cells transfected with pcDNA3 plasmids containing the coding sequences of Fas or FasL from either C57BL/6J or SEG/Pas were previously assayed for mRNA expression through quantitative real-time RT-PCR to ensure proper functionality of the constructs and the production of similar amounts of transcripts. Twenty-four hours after transfection, 5 × 103 cells transfected with a receptor were combined in 96-well microplates with a similar number of cells transfected with a ligand in the four possible combinations: (a) C57BL/6J-Fas versus C57BL/6J-FasL, (b) SEG/Pas-Fas versus SEG/Pas-FasL, (c) C57BL/6J-Fas versus SEG/Pas-FasL, and (d) SEG/Pas-Fas versus C57BL/6J-FasL.

To establish the optimal time point, caspase-8 activity was measured in cells combined for 6, 8, and 24 h. The time course experiment described a curve in which activation and reliability was maximal at 8 h (data not shown). Eight hours after combining cells, caspase-8 activity was determined using Caspase-Glo 8 Assay (Promega) according to the manufacturer's instructions.

Caspase-Glo 8 Reagent incubation time was shown to be optimal at 30 min.

Luciferase activity was monitored using a Micro Lumat Plus Microplate Luminometer LB 96 V (Berthold).

Western blot. For detection of caspase-3 and caspase-8, HEK-293 cells transfected and combined in the same conditions as for the Caspase-Glo 8 Assay (described above) were lysed using the radioimmunoprecipitation assay cell lysis buffer as described previously (23).

On the other hand, total proteins from C57BL/6J and SEG/Pas thymuses were extracted using TriPure Reagent.

Aliquots of total cell lysates containing equivalent amounts of protein (40 μg) were separated by 15% SDS-PAGE under reducing conditions and electrotransferred to pure nitrocellulose membranes (Bio-Rad). For immunodetection of caspase-3 and caspase-8 proteins, rabbit anti-mouse polyclonal antibody H-277 against caspase-3 (Santa Cruz Biotechnology) and mouse monoclonal antibody 1C12 against caspase-8 (Cell Signaling Technology), respectively, were used with appropriate secondary antibodies coupled to horseradish peroxidase. Subsequently, the peroxidase activity was obtained using Lumi-LightPLUS Western Blotting Substrate (Roche Diagnostics).

Monoclonal anti-α-tubulin (clone DM 1A) antibody (Sigma) and monoclonal anti-β-actin (clone AC 15; Sigma) were used for loading control.

Quantification of bands was made through densitometry using the Scion Image program.

TUNEL assays. Apoptosis was determined in glass-cultured HEK-293 cells by using TUNEL staining according to the manufacturer (Roche) through microscopic observation. Additionally, murine thymuses from untreated and treated mice (1.75 or 10 Gy) sacrificed 24 h after the treatment were used to separate thymic T cells from stroma cells using a 40-μm nylon cell strainer (BD Falcon). A fraction of the thymic T cells resuspended in 1× PBS were used for TUNEL assay. Cells were fixed in 3.7% paraformaldehyde (Roche) and permeabilized in 0.1% Triton X-100 sodium citrate and then stained in suspension with TUNEL following the manufacturer's guidelines. dUTP labeling of DNA strand breaks was visualized on an Olympus BX61 fluorescence microscope. The percentage of apoptotic thymic T cells (TUNEL positive) was determined through flow cytometry by analyzing 15,000 events per individual.

Statistical methods. The Levene's test was used to test for homogeneity of variances. Statistical significances were determined using a one-way ANOVA with a Tukey comparison post test. All statistical tests were carried out with the Statistical Package for the Social Sciences software (version 12.0; SPSS, Inc.).

Transcriptional expression analysis of FasL in the thymus. Because the existence of functional polymorphisms at the promoter of Fas receptor in inbred mouse strains exhibiting differences in genetic susceptibility to γ-radiation–induced T-cell lymphoblastic lymphomas had been well documented (17), the initial step was to analyze the transcriptional expression of the Fas ligand using quantitative real-time RT-PCR as it provides high sensitivity for studying transcriptional expression. In the present work, we focused on two mouse strains (C57BL/6J and SEG/Pas) that do not develop spontaneous tumors but exhibit substantial differences in genetic susceptibility to γ-radiation–induced T-cell lymphoblastic lymphomas (18, 20).

A comparison between control thymuses from each strain revealed significant differences, the most susceptible strain (C57BL/6J) being the one that exhibited the highest basal expression and the most resistant strain (SEG/Pas) being the one that showed the lowest values (Fig. 1A). When mice were treated with a single lethal dose of γ-radiation, FasL expression experienced a striking increase compared with the basal levels, but in this case, the differences between the two strains were inverted. In this case, it was SEG/Pas, the strain evidencing the highest levels, whereas C57BL/6J exhibited the lowest ones. Finally, we monitored mice treated with four sublethal doses and found that, in the thymus lymphoma–bearing mice (TL-bearing), FasL expression in T-cell lymphoblastic lymphomas decreased substantially in relation to the control group, whereas in the resistant mice (TL-free) FasL expression in their thymuses exhibited an important increase by comparison with the control thymuses.

Figure 1.

Transcriptional expression pattern of FasL determined through quantitative real-time RT-PCR. A, normalized mean FasL transcriptional expression levels in thymuses from C57BL/6J and SEG/Pas mice. “Basal” values correspond to nonirradiated thymuses. “Lethal dose-driven” values correspond to thymuses of mice irradiated with a single dose of γ-radiation (10 Gy) and sacrificed 24 h later. “No thymic lymphoma” values correspond to thymuses of mice irradiated with four doses of 1.75 Gy, which did not develop T-cell lymphoblastic lymphoma within a 25-wk latency period. “Thymic lymphoma” values correspond to thymuses of mice irradiated with four doses of 1.75 Gy, which developed a T-cell lymphoblastic lymphoma. All data represent an average of two independent experiments, each of which was done in triplicate. B, table showing the statistical values for the data: one-way ANOVA and Tukey comparison post test. As shown in the “Significance” column, differences observed in the levels of transcriptional expression between C57BL/6J and SEG/Pas were significant in all cases, with the exception of thymic lymphomas, which showed extremely low levels of FasL expression with independence of the strain. In addition, differences between groups (within the same strain) were significant in all cases.

Figure 1.

Transcriptional expression pattern of FasL determined through quantitative real-time RT-PCR. A, normalized mean FasL transcriptional expression levels in thymuses from C57BL/6J and SEG/Pas mice. “Basal” values correspond to nonirradiated thymuses. “Lethal dose-driven” values correspond to thymuses of mice irradiated with a single dose of γ-radiation (10 Gy) and sacrificed 24 h later. “No thymic lymphoma” values correspond to thymuses of mice irradiated with four doses of 1.75 Gy, which did not develop T-cell lymphoblastic lymphoma within a 25-wk latency period. “Thymic lymphoma” values correspond to thymuses of mice irradiated with four doses of 1.75 Gy, which developed a T-cell lymphoblastic lymphoma. All data represent an average of two independent experiments, each of which was done in triplicate. B, table showing the statistical values for the data: one-way ANOVA and Tukey comparison post test. As shown in the “Significance” column, differences observed in the levels of transcriptional expression between C57BL/6J and SEG/Pas were significant in all cases, with the exception of thymic lymphomas, which showed extremely low levels of FasL expression with independence of the strain. In addition, differences between groups (within the same strain) were significant in all cases.

Close modal

Polymorphisms at the FasL promoter regions. To determine whether differences in the levels of expression of FasL could be attributed to nucleotide changes in the promoter regions, we determined the spectrum of sequence variation of the FasL promoters from the two strains [data deposited in Genbank accession nos. DQ846744 (C57BL/6J) and DQ846746 (SEG/Pas)] and compared our results with the previously published sequence of M. musculus (Genbank accession no. AF045739; ref. 21). A considerable number of nucleotide changes were detected among the strains that allow for their classification into two distinct haplotypes (Fig. 2). Probably, the most important changes are three nucleotide substitutions affecting transcription factor binding sites: two at NF-E2 (−427 C/T and −430 A/G) and one at G6-Factor (−578 T/C).

Figure 2.

Comparison between the FasL promoter sequences of C57BL/6J and SEG/Pas mice. Numbers on the left, base pair position from the translational start site. Dots, sequence identity. Polymorphisms are indicated by the corresponding nucleotide substitution or deletion. Transcription factor binding sites are underlined (and also shaded and in italics when overlapping) based on proved and potential sites described previously (21, 4549). Asterisk, polymorphisms located in transcription factor binding sites.

Figure 2.

Comparison between the FasL promoter sequences of C57BL/6J and SEG/Pas mice. Numbers on the left, base pair position from the translational start site. Dots, sequence identity. Polymorphisms are indicated by the corresponding nucleotide substitution or deletion. Transcription factor binding sites are underlined (and also shaded and in italics when overlapping) based on proved and potential sites described previously (21, 4549). Asterisk, polymorphisms located in transcription factor binding sites.

Close modal

Sequence variants at the FasL promoters involve functional polymorphisms. The functionality of those variants (haplotypes) was investigated through luciferase assays using both phorbol 12-myristate 13-acetate (PMA)/ionophore and γ-irradiation activation of Jurkat T cells. Following a previous report (17), a treatment of 10 Gy was selected as an adequate dose to induce apoptosis (24).

SEG/Pas FasL promoter drove luciferase transcription to a significantly higher extent than C57BL/6J FasL promoter both in PMA/ionophore-activated and γ-radiation–activated Jurkat T cells (Fig. 3A).

Figure 3.

FasL promoter activity in activated Jurkat T cells and correlation with the level of expression. A, normalized FasL promoter activity is calculated as the ratio (normalized reporter activity of each construction/normalized reporter activity of pGL2-NA). Normalized reporter activity is calculated as the ratio (firefly luciferase value/Renilla luciferase value). pGL2-NA, “non-activated”: empty pGL2-Basic vector transfected in untreated Jurkat T cells. Jurkat T cells were treated either with PMA/ionophore or with γ-radiation; in both cases, a one-way ANOVA test revealed significant differences between the groups (P = 0.01 and 0.004, respectively). The validity of these results is supported by the fact that, in all cases, promoter activities exhibited >2-fold increase over the basic promoterless vector (50). The data represent a combined average of three different assays. B, correlation between FasL promoter activity and transcriptional expression is shown by referring data to those of C57BL/6J. FasL transcriptional expression levels correspond to thymuses of mice irradiated with 10 Gy and analyzed 24 h later. Correlation coefficients with transcriptional expression are r = 0.9107 for PMA/ionophore-activated Jurkat T cells and r = 0.9247 for γ-radiation.

Figure 3.

FasL promoter activity in activated Jurkat T cells and correlation with the level of expression. A, normalized FasL promoter activity is calculated as the ratio (normalized reporter activity of each construction/normalized reporter activity of pGL2-NA). Normalized reporter activity is calculated as the ratio (firefly luciferase value/Renilla luciferase value). pGL2-NA, “non-activated”: empty pGL2-Basic vector transfected in untreated Jurkat T cells. Jurkat T cells were treated either with PMA/ionophore or with γ-radiation; in both cases, a one-way ANOVA test revealed significant differences between the groups (P = 0.01 and 0.004, respectively). The validity of these results is supported by the fact that, in all cases, promoter activities exhibited >2-fold increase over the basic promoterless vector (50). The data represent a combined average of three different assays. B, correlation between FasL promoter activity and transcriptional expression is shown by referring data to those of C57BL/6J. FasL transcriptional expression levels correspond to thymuses of mice irradiated with 10 Gy and analyzed 24 h later. Correlation coefficients with transcriptional expression are r = 0.9107 for PMA/ionophore-activated Jurkat T cells and r = 0.9247 for γ-radiation.

Close modal

The normalized reporter activity for each promoter (luciferase value/Renilla value) referred to that of C57BL/6J was also shown to enhance these comparisons (Fig. 3B). Interestingly, there exists a significant correlation between the promoter activities and the levels of transcriptional FasL expression when mice were subjected to a single lethal dose of 10 Gy of γ-radiation (the same that was used to activate apoptosis in Jurkat T cells). Consequently, it is reasonable to assume that some of the functional polymorphisms in the FasL promoter region could account for differential promoter activities, and thus for the specific levels of mRNA expression existing in both strains, as evidenced by a significant correlation between these data.

Polymorphisms at Fas and FasL cDNA sequences. Once the existence of polymorphisms at their promoter regions was documented, we decided to compare the coding sequences of both genes to assess whether nucleotide changes exist between both strains. To this end, we sequenced the cDNAs of Fas and FasL from the C57BL/6J and SEG/Pas strains [data deposited in Genbank accession nos. DQ846748 (C57BL/6J Fas), DQ846749 (SEG/Pas Fas), and DQ846747 (SEG/Pas FasL)] and compared them with those previously published, namely that of M. musculus for Fas (Genbank accession no. NM_007987) and that of C57BL/6J for FasL (Genbank accession no. NM_010177).

By using C57BL/6J as a reference, we found a total of 24 nucleotide changes in the coding sequence of Fas from SEG/Pas. Only a minority of them (i.e., seven) did not involve the amino acid sequence (i.e., synonymous or silent) and were thus likely nonfunctional substitutions. The vast majority represented germ-line variations with possible functional consequences (Fig. 4A).

Figure 4.

Comparison between amino acid sequences of Fas and FasL in C57BL/6J and SEG/Pas mice. Numbers on the left, amino acid position from the initiation codon. Dots, sequence identity. Polymorphisms are indicated by the corresponding amino acid substitution. A, Fas signal peptide and transmembrane region are shaded to discriminate the extracellular and the intracellular regions, both in white. The extracellular region consists of three CRDs; CRD1 and CRD3 are underlined to distinguish them from CRD2. The intracellular region contains the death domain (underlined), which consists of several subdomains, among them, α2 (bolded) and α3 (boxed and italicized) subdomains. Twelve nucleotide substitutions involving amino acidic changes were recorded in the extracellular domain (P58L, Q60S, K63E, M71T, T75K, T77I, A79T, P80S, N91K, L117P, P134S, and A143T) and 5 for the intracellular domain (R209H, M255I, S258N, K312N, and T317I). B, FasL transmembrane region is shaded to distinguish the extracellular and the intracellular regions, both in white. Two nucleotide substitutions affect the amino acid sequence of the intracellular domain (A24T and C35S) and four affect the extracellular one (P141L, G178S, T184A, and E218G).

Figure 4.

Comparison between amino acid sequences of Fas and FasL in C57BL/6J and SEG/Pas mice. Numbers on the left, amino acid position from the initiation codon. Dots, sequence identity. Polymorphisms are indicated by the corresponding amino acid substitution. A, Fas signal peptide and transmembrane region are shaded to discriminate the extracellular and the intracellular regions, both in white. The extracellular region consists of three CRDs; CRD1 and CRD3 are underlined to distinguish them from CRD2. The intracellular region contains the death domain (underlined), which consists of several subdomains, among them, α2 (bolded) and α3 (boxed and italicized) subdomains. Twelve nucleotide substitutions involving amino acidic changes were recorded in the extracellular domain (P58L, Q60S, K63E, M71T, T75K, T77I, A79T, P80S, N91K, L117P, P134S, and A143T) and 5 for the intracellular domain (R209H, M255I, S258N, K312N, and T317I). B, FasL transmembrane region is shaded to distinguish the extracellular and the intracellular regions, both in white. Two nucleotide substitutions affect the amino acid sequence of the intracellular domain (A24T and C35S) and four affect the extracellular one (P141L, G178S, T184A, and E218G).

Close modal

For FasL, we found a total of 12 nucleotide changes, 6 of them involving the change of amino acid (Fig. 4B).

Sequence variants at Fas and FasL cDNAs involve functional polymorphisms. To determine whether these nucleotide changes could affect the ability of the Fas/FasL system to induce apoptosis, the levels of induced apoptosis through in vitro experiments based on the confrontation of human HEK-293 cells transfected with cDNAs of either receptors or ligands from the C57BL/6J and SEG/Pas strains were investigated. The functionality of these changes was determined by measuring caspase-8 activity through measurement of the luciferase-linked activity (Caspase-Glo 8 Assay). We chose HEK-293 as the recipient cells because these do not express Fas and express a much reduced amount of the Fas ligand compared with the exogenous expression (data not shown).

Our results revealed significant differences in the functionality of the Fas-FasL system depending on the allelic variants involved (Fig. 5A), the SEG/Pas system being substantially more active than that of C57BL/6J in inducing apoptosis. In the heterologous combinations, mean values were intermediate.

Figure 5.

Apoptotic activity of C57BL/6J- and SEG/Pas-derived Fas/FasL systems in vitro. A, Caspase-Glo 8 Assay measures the amount of active caspase-8 in a luciferase-linked manner. HEK-293 cells transiently transfected with C57BL/6J- or SEG/Pas-derived Fas and FasL were combined and analyzed 8 h later for caspase-8 activity. Normalized caspase-8 activity in each system is expressed as the ratio [(mean relative luminescent unit value for each system − Blank) / (mean relative luminescent unit value for C57BL/6J − Blank)]. Blank: mean relative luminescent unit of medium without cells. SDs and statistical analysis are shown in the main text. All data represent an average of three independent experiments, each one carried out in triplicate. A one-way ANOVA and a Tukey comparison post test evidenced significant differences among the groups (P < 0.01), except for the two heterologous combinations. B, Western blots for caspase-8 and caspase-3 activation using cell extracts from the same combinations used in (A) for the Caspase-Glo 8 Assay. Value under each band, expression of cleaved caspase, normalized to α-tubulin and referred to that of C57BL/6J Fas-C57BL/6J FasL. C, quantification of TUNEL-positive cells. The number of TUNEL-positive cells was counted from a total number of at least 250 cells, estimated through optical microscopy, and represented as the percentage of positive cells as referred to the total.

Figure 5.

Apoptotic activity of C57BL/6J- and SEG/Pas-derived Fas/FasL systems in vitro. A, Caspase-Glo 8 Assay measures the amount of active caspase-8 in a luciferase-linked manner. HEK-293 cells transiently transfected with C57BL/6J- or SEG/Pas-derived Fas and FasL were combined and analyzed 8 h later for caspase-8 activity. Normalized caspase-8 activity in each system is expressed as the ratio [(mean relative luminescent unit value for each system − Blank) / (mean relative luminescent unit value for C57BL/6J − Blank)]. Blank: mean relative luminescent unit of medium without cells. SDs and statistical analysis are shown in the main text. All data represent an average of three independent experiments, each one carried out in triplicate. A one-way ANOVA and a Tukey comparison post test evidenced significant differences among the groups (P < 0.01), except for the two heterologous combinations. B, Western blots for caspase-8 and caspase-3 activation using cell extracts from the same combinations used in (A) for the Caspase-Glo 8 Assay. Value under each band, expression of cleaved caspase, normalized to α-tubulin and referred to that of C57BL/6J Fas-C57BL/6J FasL. C, quantification of TUNEL-positive cells. The number of TUNEL-positive cells was counted from a total number of at least 250 cells, estimated through optical microscopy, and represented as the percentage of positive cells as referred to the total.

Close modal

Because caspase-8 activation should occur through the cleavage of the procaspase form (57 kDa) into an intermediate form of 43 kDa, we likewise measured caspase-8 activation through Western blot. Consistent with data on caspase-8 as determined by the Caspase-Glo 8 method, we found that the amount of cleaved caspase-8 protein is significantly higher in the SEG/Pas Fas-SEG/Pas FasL system by comparison with the C57BL/6J Fas-C57BL/6J FasL system. Once again, the heterologous systems exhibited intermediate values (Fig. 5B).

Caspase-8, which is specifically activated in the Fas/FasL extrinsic pathway (25, 26), subsequently cleaves and activates the effector caspase-3. We therefore also monitored caspase-3 activation through Western blot by determining procaspase-3 (32 kDa) and the active p17 subunit. Our results showed that the SEG/Pas system exhibited a higher level of the p17 subunit by comparison with the C57BL/6J system, whereas the heterologous systems exhibited intermediate amounts of cleaved caspase-3 (Fig. 5B).

Once it has been shown that the allelic variants of the Fas/FasL system determine specific patterns of caspase-8 and caspase-3 activation, we wanted to prove whether these have a global effect in apoptosis. To this end, we analyzed apoptosis using a TUNEL assay in HEK-293 cells transfected with either C57BL/6J-derived or SEG/Pas-derived Fas/FasL system. The results (Fig. 5C; Supplementary Data 1) show that the SEG/Pas system drives cell apoptosis to a significantly higher extent than the C57BL/6J system.

Additionally, caspase-8 activation was measured through Western blot in thymuses from C57BL/6J and SEG/Pas mice both before and after their 10 Gy γ-irradiation. Not only the amount of cleaved caspase-8 protein is significantly higher in SEG/Pas mice but also the total amount of protein increases after γ-irradiation to a higher extent for SEG/Pas (Fig. 6A). Furthermore, we analyzed the apoptotic levels in thymus-derived T cells from both C57BL/6J and SEG/Pas mice 24 h after irradiation with either a single sublethal dose of 1.75 Gy or a single lethal dose of 10 Gy. This was accomplished by counting the number of TUNEL-positive (apoptotic) cells. The results evidence a striking difference in the percentage of apoptotic T cells (Fig. 6B). These data confirm that the SEG/Pas-derived Fas/FasL system is far more efficient than the C57BL/6J-derived system and that it leads to a significant increase in γ-radiation–induced apoptosis of thymic T cells.

Figure 6.

Apoptotic activity of thymic T cells from C57BL/6J and SEG/Pas mice. A, Western blot for caspase-8 activation using protein extracts from C57BL/6J and SEG/Pas thymuses, both untreated (CTRL) and subjected to a single dose of 10 Gy and analyzed 24 h later. Value under each band, expression of cleaved caspase, normalized to β-actin and referred to that of untreated C57BL/6J. In addition, the total amounts of the protein (procaspase plus cleaved form), normalized to β-actin and referred to that of untreated C57BL/6J, are as follows: C57BL/6J CTRL, 1.00; C57BL/6J 10 Gy, 2.06; SEG/Pas CTRL, 1.09; and SEG/Pas 10 Gy, 3.13. B, quantification of TUNEL-positive cells. The number of TUNEL-positive cells was determined through flow cytometry and represented as the percentage of positive cells as referred to the total.

Figure 6.

Apoptotic activity of thymic T cells from C57BL/6J and SEG/Pas mice. A, Western blot for caspase-8 activation using protein extracts from C57BL/6J and SEG/Pas thymuses, both untreated (CTRL) and subjected to a single dose of 10 Gy and analyzed 24 h later. Value under each band, expression of cleaved caspase, normalized to β-actin and referred to that of untreated C57BL/6J. In addition, the total amounts of the protein (procaspase plus cleaved form), normalized to β-actin and referred to that of untreated C57BL/6J, are as follows: C57BL/6J CTRL, 1.00; C57BL/6J 10 Gy, 2.06; SEG/Pas CTRL, 1.09; and SEG/Pas 10 Gy, 3.13. B, quantification of TUNEL-positive cells. The number of TUNEL-positive cells was determined through flow cytometry and represented as the percentage of positive cells as referred to the total.

Close modal

Recently, our group reported that Fas expression is scarcely perceptible in murine γ-radiation–induced T-cell lymphoblastic lymphomas (16) and suggested the involvement of some promoter variants in modulating the resistance/susceptibility of murine strains to this kind of tumors (17). Because Fas is a receptor molecule acting in conjunction with its ligand (FasL), the first goal of the present study was to assess whether FasL is also involved in this susceptibility. To this end, the in vivo pattern of transcriptional expression of this gene was analyzed in two mouse strains (C57BL/6J and SEG/Pas) exhibiting extreme differences in genetic susceptibility to γ-radiation–induced T-cell lymphoblastic lymphomas (Fig. 1). Our finding that a striking early overexpression occurred in vivo after treatment with γ-irradiation (with independence of the strain), together with the aforementioned data, suggests the existence of a concomitant action by both receptor and ligand (i.e., the Fas/FasL system) to prevent cellular transformation. Striking was the fact that a significant increase in the levels of FasL expression was also found later after the period of latency in mice that resisted tumor induction. Such an increase might be attributable to some unknown epigenetic mechanism and remarks the importance of a putative Fas/FasL barrier in maintaining a pervasive phenotype of resistance long after the time of irradiations. On the other hand, because FasL expression is practically absent in T-cell lymphoblastic lymphomas, it would be reasonable to think that such a reduction can break the proposed Fas/FasL-defensive wall down and that this event could be an essential step to attain a tumor phenotype. Furthermore, because both of these strains share a common level of FasL expression in thymic lymphomas, the hypothesis of a “tumor threshold” that we proposed for Fas now becomes also plausible for FasL. Taken together, these data strongly support a role for both Fas and FasL as T-cell lymphoma suppressor genes.

When the Fas/FasL system does not operate properly, many diseases arise, including cancer, acquired immune deficiency syndrome, and autoimmune disorders (1, 6, 2729). Specifically, heterozygous germ-line Fas point mutations have been reported to cause autoimmune lymphoproliferative syndrome with defective apoptosis and elevated risk to develop lymphomas, basically of B-cell types (27, 30). It has been proposed that induction of apoptosis by this system requires two wild-type alleles for Fas and that a proliferating pathway involving nuclear factor-κB (NF-κB) can be fully activated in cells expressing the mutant and wild-type Fas alleles (30). However, no one has yet studied the levels of expression of the wild-type alleles to show whether cancer predisposition is really due to a decrease in their levels of expression or, alternatively, if the presence of the mutated alleles is also necessary. Our results suggest a scenario where decreased levels of wild-type alleles from both Fas and FasL may promote tumor susceptibility, probably by favoring the pathway of NF-κB in detriment of apoptosis as we mentioned above (30). In addition, the development of T-cell lymphoblastic lymphomas would require that the transcriptional levels of these genes were lowered beyond a tumor threshold.

The concomitant increase (after a γ-radiation treatment) or decrease (in T-cell lymphoblastic lymphomas) in the expression of both Fas and FasL could be due to these genes sharing some common regulatory elements at their promoter regions (most importantly NF-κB and NF-AT). However, such parallel behavior contradicts to a certain degree what has been published for other tumors. Several authors have reported that solid tumors usually express elevated levels of FasL but reduced levels of Fas receptor (31, 32). Up-regulation of FasL has been linked to some advantage for tumor cells through the triggering of apoptosis-independent pathways, resulting in up-regulation of antiapoptotic and tumorigenic genes and/or an increase in motility and invasiveness, particularly for apoptosis-resistant tumor cells (33). On the contrary, our findings indicate that T-cell lymphoblastic lymphomagenesis does not require increased levels of FasL to proliferate but instead the simultaneous down-regulation of both Fas and FasL to avoid excessive apoptosis.

To determine whether FasL might act as a germ-line modulator of resistance/susceptibility, as we proposed for Fas (17), the levels of expression of FasL in control and γ-radiation–treated mice from both strains were examined. We had hypothesized that in case decreasing levels of FasL could contribute to T-cell lymphoma development, the different capability to produce FasL showed by C57BL/6J and SEG/Pas mice might contribute to the extreme differences in susceptibility exhibited by those strains.

Interestingly, although the most susceptible strain (C57BL/6J) exhibits the highest basal level of FasL expression, a single dose of γ-irradiation is able to switch this pattern in a mere of 24 h, generating a 7.5-fold increase for the SEG/Pas strain, whereas the C57BL/6J strain only exhibits a 3.6-fold increase. Thus, it seems reasonable to assume that T-cell lymphoblastic lymphomas are more frequent in C57BL/6J, which exhibits a lower defensive radiation response. However, the magnitude of transcriptional and luciferase differences might not be sufficient to produce the remarkable difference in lymphoma incidence. Thus, the implication of another regulatory gene to eliminate the high expression induced by irradiation in C57BL/6J cannot be ruled out. Furthermore, the shared contribution of both the Fas receptor and the Fas ligand to the tumor threshold could be a novel yet attractive idea.

Because several reports have suggested an association between polymorphisms at the FasL promoter and the risk of developing several types of diseases, including some solid cancers and autoimmune diseases in humans (3437), we decided to analyze the FasL promoter sequence in both the C57BL/6J and SEG/Pas strains. Many nucleotide changes along the entire promoter region (Fig. 2), some affecting the transcription factor binding sites, such as the G6-Factor and NF-E2, were found. Remarkably, the two resultant haplotypes exhibited significant differences of functionality in luciferase assays after both PMA/ionophore and γ-radiation induction were administrated (Fig. 3A), suggesting that the reported FasL promoter differences may be contributing to the pattern of differential expression of FasL between these strains (Fig. 3B).

Once the importance of the functional polymorphisms at the promoter regions was shown, we analyzed the contribution of the coding sequences to the biological activity of the Fas/FasL system. A comparative analysis revealed that cDNAs from both Fas and FasL exhibited several nucleotide changes; some of which involved a change of amino acid (Fig. 4).

For Fas (Fig. 4A), there were changes located in the tumor necrosis factor receptor–like extracellular cysteine-rich domain (CRD) 1, as well as in CRD2, in CRD3, and in the intracellular region. With regard to the extracellular region, because CRD1 is involved in the preligand assembly domain formation (3840) and the interaction between Fas and FasL is known to occur through CRD2 and CRD3, it is presumed that changes detected in these domains can constitute functional polymorphisms (41, 42). As for the intracellular region, the five amino acidic changes found (two of them located inside the α3 subdomain of death domain) could alter the interaction of Fas with FADD to form the DISC and trigger the death signal (43). Such assumption is reinforced by an association that has been found between a point mutation in Fas death domain and an autoimmune lymphoproliferative syndrome reported in humans, for both T-cell lymphoma and Hodgkin's disease (44).

As for FasL (Fig. 4B), two of the four amino acid changes in the extracellular region (T184A and E218G) have been already reported as being relevant for the effectiveness of the Fas/FasL system because the D-E loop, where both polymorphisms are located, is directly involved in ligand-receptor contact (19). Therefore, the substitution G218 to E218 may lead to a conformational change of this loop, or to the partial loss of the contact with the receptor, which would result in a reduced binding affinity and cytotoxic activity of the E218 variant. Because C57BL/6J bears the E218 variant whereas SEG/Pas exhibits the G218 variant, we believe that this polymorphism could be contributing to a less effective interaction between the receptor and its ligand, thus subsequently to a lower activity of the Fas/FasL system from C57BL/6J mice by comparison with that of SEG/Pas strain.

To determine whether these amino acid changes represent functional polymorphisms capable of modulating the biological activity of the Fas/FasL system, the ability of the system of each strain to induce the death signal, through measuring caspase-8 activation in transfected HEK-293 cells, was tested (Fig. 5A). Our data show that the SEG/Pas-derived Fas/FasL system is likewise more efficient than that derived from C57BL/6J, with a striking difference in the activation of caspase-8 of ∼112%, evidencing the physiologic relevance of the aforementioned amino acid changes. This hypothesis supports the idea that SEG/Pas mice bear a more effective Fas-dependent apoptosis pathway than the C57BL/6J mice do. Such assumption was corroborated through Western blot against caspase-8 and caspase-3 and through measure of the apoptotic levels (Figs. 5B and C and 6) both in vitro and in vivo.

In conclusion, three important points have emerged from these analyses. The first one is that the expression of FasL increases immediately after γ-radiation treatment and is practically absent in T-cell lymphoblastic lymphomas, suggesting a role for FasL as a tumor suppressor gene. Second, the study shows that there exist significant differences in the levels of FasL expression between the thymuses of two mouse strains extremely different in terms of genetic susceptibility that could be related to functional polymorphisms at the promoter regions. The third point emerging from our research is that a significant amount of polymorphisms in the coding sequences of Fas and FasL can affect the biological activity of the Fas/FasL system, as shown in vitro by transfections of HEK-293 cells, and in vivo, in the T cells inside the thymuses of mice from the two strains.

These data support a model in which the Fas/FasL system may play a role in modulating the genetic susceptibility of mouse strains to develop T-cell lymphoblastic lymphomas.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: European Commission EC RISC-RAD contract FI6R-CT-2003-508842 (J. Santos) and Spanish Ministry of Education and Science grants SAF2003-05048 (J. Fernández-Piqueras) and SAF2006-09437 (M. Villa-Morales and J. Fernández-Piqueras). M. Villa-Morales was the recipient of a predoctoral fellowship from the Spanish Ministry of Education and Science (FPU AP2002-0294) up to December 2006.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank L. González and P. Cozar for aid with animals, S. Jiménez-Baranda for technical assistance, Drs. J.L. Guenet and X. Montagutelli for providing SEG/Pas mice, and Prof. A. Morales for critical reading of the manuscript.

1
Baumann S, Krueger A, Kirchhoff S, Krammer PH. Regulation of T cell apoptosis during the immune response.
Curr Mol Med
2002
;
2
:
257
–72.
2
Schmitz I, Krueger A, Baumann S, Schulze-Bergkamen H, Krammer PH, Kirchhoff S. An IL-2-dependent switch between CD95 signaling pathways sensitizes primary human T cells toward CD95-mediated activation-induced cell death.
J Immunol
2003
;
171
:
2930
–6.
3
Moulian N, Berrih-Aknin S. Fas/APO-1/CD95 in health and autoimmune disease: thymic and peripheral aspects.
Semin Immunol
1998
;
10
:
449
–56.
4
Siegel RM, Chan FK, Chun HJ, Lenardo MJ. The multifaceted role of Fas signaling in immune cell homeostasis and autoimmunity.
Nat Immunol
2000
;
1
:
469
–74.
5
Nagata S. Fas-induced apoptosis, and diseases caused by its abnormality.
Genes Cells
1996
;
1
:
873
–9.
6
Krammer PH. CD95's deadly mission in the immune system.
Nature
2000
;
407
:
789
–95.
7
Muschen M, Warskulat U, Beckmann MW. Defining CD95 as a tumor suppressor gene.
J Mol Med
2000
;
78
:
312
–25.
8
Rathmell JC, Thompson CB. Pathways of apoptosis in lymphocyte development, homeostasis, and disease.
Cell
2002
;
109
Suppl:
S97
–107.
9
Fleck M, Zhou T, Tatsuta T, Yang P, Wang Z, Mountz JD. Fas/Fas ligand signaling during gestational T cell development.
J Immunol
1998
;
160
:
3766
–75.
10
Kishimoto H, Surh CD, Sprent J. A role for Fas in negative selection of thymocytes in vivo.
J Exp Med
1998
;
187
:
1427
–38.
11
Kurasawa K, Hashimoto Y, Iwamoto I. Fas modulates both positive and negative selection of thymocytes.
Cell Immunol
1999
;
194
:
127
–35.
12
Schmitz I, Meyer C, Schulze-Osthoff K. CD95 ligand mediates T-cell receptor-induced apoptosis of a CD4+ CD8+ double positive thymic lymphoma.
Oncogene
2006
;
25
:
7587
–96.
13
Guerrero I, Villasante A, Corces V, Pellicer A. Activation of a c-K-ras oncogene by somatic mutation in mouse lymphomas induced by γ radiation.
Science
1984
;
225
:
1159
–62.
14
Booker JK, Reap EA, Cohen PL. Expression and function of Fas on cells damaged by γ-irradiation in B6 and B6/lpr mice.
J Immunol
1998
;
161
:
4536
–41.
15
Pinkoski MJ, Green DR. Fas ligand, death gene.
Cell Death Differ
1999
;
6
:
1174
–81.
16
Santos J, Herranz M, Fernandez M, Vaquero C, Lopez P, Fernandez-Piqueras J. Evidence of a possible epigenetic inactivation mechanism operating on a region of mouse chromosome 19 in γ-radiation-induced thymic lymphomas.
Oncogene
2001
;
20
:
2186
–9.
17
Villa-Morales M, Santos J, Fernandez-Piqueras J. Functional Fas (Cd95/Apo-1) promoter polymorphisms in inbred mouse strains exhibiting different susceptibility to γ-radiation-induced thymic lymphoma.
Oncogene
2006
;
25
:
2022
–9.
18
Santos J, Montagutelli X, Acevedo A, et al. A new locus for resistance to γ-radiation-induced thymic lymphoma identified using inter-specific consomic and inter-specific recombinant congenic strains of mice.
Oncogene
2002
;
21
:
6680
–3.
19
Kayagaki N, Yamaguchi N, Nagao F, et al. Polymorphism of murine Fas ligand that affects the biological activity.
Proc Natl Acad Sci U S A
1997
;
94
:
3914
–9.
20
Santos J, Perez de Castro I, Herranz M, Pellicer A, Fernandez-Piqueras J. Allelic losses on chromosome 4 suggest the existence of a candidate tumor suppressor gene region of about 0.6 cM in γ-radiation-induced mouse primary thymic lymphomas.
Oncogene
1996
;
12
:
669
–76.
21
Zhang J, Miranda K, Ma BY, Fine A. Molecular characterization of the mouse Fas ligand promoter in airway epithelial cells.
Biochim Biophys Acta
2000
;
1490
:
291
–301.
22
Perez-Gomez E, Eleno N, Lopez-Novoa JM, et al. Characterization of murine S-endoglin isoform and its effects on tumor development.
Oncogene
2005
;
24
:
4450
–61.
23
Quintanilla M, Ramirez JR, Perez-Gomez E, et al. Expression of the TGF-β coreceptor endoglin in epidermal keratinocytes and its dual role in multistage mouse skin carcinogenesis.
Oncogene
2003
;
22
:
5976
–85.
24
Maas K, Westfall M, Pietenpol J, Olsen NJ, Aune T. Reduced p53 in peripheral blood mononuclear cells from patients with rheumatoid arthritis is associated with loss of radiation-induced apoptosis.
Arthritis Rheum
2005
;
52
:
1047
–57.
25
Barnhart BC, Alappat EC, Peter ME. The CD95 type I/type II model.
Semin Immunol
2003
;
15
:
185
–93.
26
Scaffidi C, Fulda S, Srinivasan A, et al. Two CD95 (APO-1/Fas) signaling pathways.
EMBO J
1998
;
17
:
1675
–87.
27
Peter ME, Heufelder AE, Hengartner MO. Advances in apoptosis research.
Proc Natl Acad Sci U S A
1997
;
94
:
12736
–7.
28
Li-Weber M, Krammer PH. Function and regulation of the CD95 (APO-1/Fas) ligand in the immune system.
Semin Immunol
2003
;
15
:
145
–57.
29
Peter ME, Legembre P, Barnhart BC. Does CD95 have tumor promoting activities?
Biochim Biophys Acta
2005
;
1755
:
25
–36.
30
Legembre P, Barnhart BC, Zheng L, et al. Induction of apoptosis and activation of NF-κB by CD95 require different signalling thresholds.
EMBO Rep
2004
;
5
:
1084
–9.
31
Botti C, Buglioni S, Benevolo M, et al. Altered expression of FAS system is related to adverse clinical outcome in stage I-II breast cancer patients treated with adjuvant anthracycline-based chemotherapy.
Clin Cancer Res
2004
;
10
:
1360
–5.
32
Soubrane C, Mouawad R, Antoine EC, Verola O, Gil-Delgado M, Khayat D. A comparative study of Fas and Fas-ligand expression during melanoma progression.
Br J Dermatol
2000
;
143
:
307
–12.
33
Barnhart BC, Legembre P, Pietras E, Bubici C, Franzoso G, Peter ME. CD95 ligand induces motility and invasiveness of apoptosis-resistant tumor cells.
EMBO J
2004
;
23
:
3175
–85.
34
Zhang X, Miao X, Sun T, et al. Functional polymorphisms in cell death pathway genes FAS and FASL contribute to risk of lung cancer.
J Med Genet
2005
;
42
:
479
–84.
35
Sun T, Miao X, Zhang X, Tan W, Xiong P, Lin D. Polymorphisms of death pathway genes FAS and FASL in esophageal squamous-cell carcinoma.
J Natl Cancer Inst
2004
;
96
:
1030
–6.
36
Sun T, Zhou Y, Li H, et al. FASL −844C polymorphism is associated with increased activation-induced T cell death and risk of cervical cancer.
J Exp Med
2005
;
202
:
967
–74.
37
Wu J, Metz C, Xu X, et al. A novel polymorphic CAAT/enhancer-binding protein β element in the FasL gene promoter alters Fas ligand expression: a candidate background gene in African American systemic lupus erythematosus patients.
J Immunol
2003
;
170
:
132
–8.
38
Siegel RM, Frederiksen JK, Zacharias DA, et al. Fas preassociation required for apoptosis signaling and dominant inhibition by pathogenic mutations.
Science
2000
;
288
:
2354
–7.
39
Golstein P. Signal transduction. FasL binds preassembled Fas.
Science
2000
;
288
:
2328
–9.
40
Papoff G, Hausler P, Eramo A, et al. Identification and characterization of a ligand-independent oligomerization domain in the extracellular region of the CD95 death receptor.
J Biol Chem
1999
;
274
:
38241
–50.
41
Schneider P, Bodmer JL, Holler N, et al. Characterization of Fas (Apo-1, CD95)-Fas ligand interaction.
J Biol Chem
1997
;
272
:
18827
–33.
42
Starling GC, Kiener PA, Aruffo A, Bajorath J. Analysis of the ligand binding site in Fas (CD95) by site-directed mutagenesis and comparison with TNFR and CD40.
Biochemistry
1998
;
37
:
3723
–6.
43
Jeong EJ, Bang S, Lee TH, Park YI, Sim WS, Kim KS. The solution structure of FADD death domain. Structural basis of death domain interactions of Fas and FADD.
J Biol Chem
1999
;
274
:
16337
–42.
44
Peters AM, Kohfink B, Martin H, et al. Defective apoptosis due to a point mutation in the death domain of CD95 associated with autoimmune lymphoproliferative syndrome, T-cell lymphoma, and Hodgkin's disease.
Exp Hematol
1999
;
27
:
868
–74.
45
Crist SA, Griffith TS, Ratliff TL. Structure/function analysis of the murine CD95L promoter reveals the identification of a novel transcriptional repressor and functional CD28 response element.
J Biol Chem
2003
;
278
:
35950
–8.
46
Latinis KM, Norian LA, Eliason SL, Koretzky GA. Two NFAT transcription factor binding sites participate in the regulation of CD95 (Fas) ligand expression in activated human T cells.
J Biol Chem
1997
;
272
:
31427
–34.
47
Matsui K, Fine A, Zhu B, Marshak-Rothstein A, Ju ST. Identification of two NF-κB sites in mouse CD95 ligand (Fas ligand) promoter: functional analysis in T cell hybridoma.
J Immunol
1998
;
161
:
3469
–73.
48
Matsui K, Xiao S, Fine A, Ju ST. Role of activator protein-1 in TCR-mediated regulation of the murine fasl promoter.
J Immunol
2000
;
164
:
3002
–8.
49
McClure RF, Heppelmann CJ, Paya CV. Constitutive Fas ligand gene transcription in Sertoli cells is regulated by Sp1.
J Biol Chem
1999
;
274
:
7756
–62.
50
Hoogendoorn B, Coleman SL, Guy CA, et al. Functional analysis of human promoter polymorphisms.
Hum Mol Genet
2003
;
12
:
2249
–54.

Supplementary data